#!/usr/bin/perl -w
use strict;

### execute script and save content in $output
my $output = `./2.pl argument.txt`;
my $filename = "./output.txt"; 

### write output to file $filename 
open (OUTFILE, ">$filename") or die "Can't open $filename: $!";
print OUTFILE $output;
close OUTFILE;
[download]

open (OUTFILE, ">/path/to/file.txt") or die "Can't open file for text 
+output!\n";
print OUTFILE 'Some text you want to print to the file';
close OUTFILE;
[download]

While you are quite right that the exec function in perl can be used to execute a command, it is seldom used by itself. Usually it is preceded by the fork function call.

To execute a program and fetch it's output, I would rather suggest either using the qx() construct or doing it by hand, open(F, "program|").

Autark.

If the script is on the same server as the original script then you can run it using system.

If, however, as your example implies you are running a CGI script on another server, then you'll need to use something like the LWP module (available on the CPAN).

"Perl makes the fun jobs fun
and the boring jobs bearable" - me

assuming 1.pl is a CGI script, called via a browser with http://something/cgi-bin/1.pl?file=filename.ext

you can get to the filename being passed in 1.pl simply by doing something like this:

use CGI;
my $query = new CGI;
my $filename = $query->param("file");  # $filename now contains filena
+me.ext
[download]

Now you have the filename in the variable $file

English is not my first language, so I am having some trouble explaining what I want to do. :o)
You are correct in your assumption that 1.pl is usually call via a browser using:
http://www.server.com/cgi-bin/1.pl?filename.txt
But it is not the filename.txt I want.
I want the result of the above url/execution written to a file, using a second script, 2.pl

The feasibility of this depends totally on the nature of the CGI script you're wanting to call by hand. CGI scripts by their nature expect to be called from a web server, expect their arguments to provided per CGI spec, and will provide CGI output. The script.pl?argument construct is a web/CGI thing, but since you're trying to call your script via standard system() methods, CGI methods don't apply. Fortunately, the "GET" method you're using makes it easy:

system("./1.pl filename.txt");      # to stdout, or:
my $output = `./1.pl filename.txt`  # in $output
[download]

A security note: If, instead of using "filename.txt", you use a user-provided variable, make SURE your script uses taint-checking (-T argument, see perlsec) and un-taint the variable smartly before attempting to use it, or someone can easily put in a filename of "/etc/passwd" or "some.file;rm -rf /etc/httpd|", for example.

Since the script is a CGI script, its output will contain CGI headers. If this is OK, great. Otherwise, you will need to strip them out.

Perhaps you should consider an ASP technique for running a CGI script "within" the context of a page (with SSI this is called a "virtual include").

Hope this helps.

My intent was to quickly show how to get the passed parameter easily using perl's CGI.pm module (since you indicated your experience was with ASP). Wasn't attempting to address what would be done with the data.

Seems that your 2.pl could use LWP to call the above file, storing the results in a variable/array.

or you could execute 1.pl from the command line using CGI.pm

perl 1.pl file=filename.txt > outfile
[download]

which could also be done via a 2.pl program, shell script or .bat file.

Hope these bits and pieces are helpful